Commentary: significant improvements in weight, body mass index and waist circumference for participants at the end of a 10-week health promotion intervention for people with mental disorders were not maintained at the six-month follow-up.

[Extract] The physiological, psycho-social and cognitive benefits of regular physical activity (PA) have been well documented for the general population. Less evidence is available on the benefits of exercise for individuals with mental disorders (Alexandratos, Barnett & Thomas, 2012). This paper goes someway to address this issue. The use of a cluster preference randomised controlled trial design was an appropriate choice to reduce the risk of contamination bias. However, the addition of preference design may have introduced volunteer and measurement bias. The small but significant results in primary outcomes may be due to intervention limitations